five

open-llm-leaderboard-old/details_TwT-6__open_llm_leaderboard_demo

收藏
Hugging Face2024-04-19 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_TwT-6__open_llm_leaderboard_demo
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在模型 TwT-6/open_llm_leaderboard_demo 在 Open LLM Leaderboard 上的评估运行期间自动创建的。数据集由 63 个配置组成,每个配置对应一个被评估的任务。数据集是从 2 次运行中创建的,每次运行在每个配置中表示为特定的分割,分割名称使用运行的时间戳。train 分割始终指向最新的结果。一个名为 results 的额外配置存储了所有运行的聚合结果,用于计算和显示 Open LLM Leaderboard 上的聚合指标。README 还提供了如何使用 Python 中的 datasets 库加载运行细节的示例。

该数据集是在模型 TwT-6/open_llm_leaderboard_demo 在 Open LLM Leaderboard 上的评估运行期间自动创建的。数据集由 63 个配置组成,每个配置对应一个被评估的任务。数据集是从 2 次运行中创建的,每次运行在每个配置中表示为特定的分割,分割名称使用运行的时间戳。train 分割始终指向最新的结果。一个名为 results 的额外配置存储了所有运行的聚合结果,用于计算和显示 Open LLM Leaderboard 上的聚合指标。README 还提供了如何使用 Python 中的 datasets 库加载运行细节的示例。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

数据集创建

数据集结构

  • 配置数量:数据集包含 63 个配置,每个配置对应一个评估任务。
  • 运行次数:数据集从 2 次运行中创建。每个运行在每个配置中作为一个特定的分割存在,分割名称使用运行的时间戳。
  • 最新结果:"train" 分割始终指向最新的结果。
  • 聚合结果:一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示在 Open LLM Leaderboard 上的聚合指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_TwT-6__open_llm_leaderboard_demo", "harness_winogrande_5", split="train")

最新结果

  • 最新结果时间戳:2024-04-19T22:13:16.319405
  • 结果示例: python { "all": { "acc": 0.6840706308355067, "acc_stderr": 0.031694687760243416, "acc_norm": 0.6874422634985915, "acc_norm_stderr": 0.032324412776318714, "mc1": 0.41003671970624234, "mc1_stderr": 0.017217844717449325, "mc2": 0.5821442446195572, "mc2_stderr": 0.0148654708562592 }, "harness|arc:challenge|25": { "acc": 0.5298634812286689, "acc_stderr": 0.014585305840007105, "acc_norm": 0.5802047781569966, "acc_norm_stderr": 0.01442218122630303 }, "harness|hellaswag|10": { "acc": 0.6195976897032464, "acc_stderr": 0.004844935327599205, "acc_norm": 0.8158733320055765, "acc_norm_stderr": 0.0038679533932836465 }, # 其他任务的结果... }

配置详情

  • 配置名称:harness_arc_challenge_25

    • 数据文件
      • 分割:2024_04_16T10_37_38.522971
        • 路径:**/details_harness|arc:challenge|25_2024-04-16T10-37-38.522971.parquet
      • 分割:2024_04_19T22_13_16.319405
        • 路径:**/details_harness|arc:challenge|25_2024-04-19T22-13-16.319405.parquet
      • 分割:latest
        • 路径:**/details_harness|arc:challenge|25_2024-04-19T22-13-16.319405.parquet
  • 配置名称:harness_gsm8k_5

    • 数据文件
      • 分割:2024_04_16T10_37_38.522971
        • 路径:**/details_harness|gsm8k|5_2024-04-16T10-37-38.522971.parquet
      • 分割:2024_04_19T22_13_16.319405
        • 路径:**/details_harness|gsm8k|5_2024-04-19T22-13-16.319405.parquet
      • 分割:latest
        • 路径:**/details_harness|gsm8k|5_2024-04-19T22-13-16.319405.parquet
  • 配置名称:harness_hellaswag_10

    • 数据文件
      • 分割:2024_04_16T10_37_38.522971
        • 路径:**/details_harness|hellaswag|10_2024-04-16T10-37-38.522971.parquet
      • 分割:2024_04_19T22_13_16.319405
        • 路径:**/details_harness|hellaswag|10_2024-04-19T22-13-16.319405.parquet
      • 分割:latest
        • 路径:**/details_harness|hellaswag|10_2024-04-19T22-13-16.319405.parquet
  • 配置名称:harness_hendrycksTest_5

    • 数据文件
      • 分割:2024_04_16T10_37_38.522971
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-international_law|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-management|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-marketing|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2024-04-16T10-37-38.522971.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2024-04-16T10-37-38.522971.parquet
          • `**/details_harness|hendrycksTest-philosophy|5_2024-04-16T10-37-38.5
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作